In svenhalvorson/qualtreats: Create tables of a Qualtrics surveys from the API

qualtables

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = '#>',
  fig.path = 'man/figures/README-',
  out.width = '100%',
  warning = FALSE,
  message = FALSE
)

This is under development. If you find bugs or wish to collaborate, please contact me at [e-mail]

Huge thank you to authors of qualtRics. Appreciate learn/steal-ing your httr code

qualtables is a library that can retrieve data about Qualtrics surveys extracted from the API. You can generate tables that describe the questions, blocks, choices, and exported columns of a survey. This is particularly useful for checking that the survey is configured correctly or comparing surveys.

Install it like this:

devtools::install_github('svenhalvorson/qualtables')

These two environment variables (should be accessible by Sys.getenv) need to be set in the same way that the qualtRics library uses them:

1. QUALTRICS_API_KEY: 'your_api_key'
2. QUALTRICS_BASE_URL: 'oregon.ca1.qualtrics.com' (depends on organization)

Flatten a survey

The function flatten_survey creates a semi-curated version of the survey definitions from this API call. Use a survey ID (can be found in the URL) to flatten the survey into three tables:

survey_id = 'SV_bg4hf9VdW9CwmiO'

survey_flat = qualtables::flatten_survey(survey_id)

Here are a few of the columns in the questions table:

survey_flat[['questions']] |> 
  dplyr::select(question_id, question_description, question_type, question_selector) |>
  head()

And the choices table:

head(survey_flat[['choices']])

There is also a blocks table which is typically less interesting (but still useful).

Simplify question types

The next function, simplify_qtypes, makes a list of all the questions and assigns some flags to them. It will tell you if a question is a matrix, side-by-side (sbs), or is part of a loop-and-merge. It categorizes the questions by question_style. This can accept either the survey_id or a flattened survey (faster).

survey_qtypes = qualtables::simplify_qtypes(survey_flat = survey_flat)
head(survey_qtypes)

Get a column map

When working with survey responses, I found it useful to know the associations between what Qualtrics exports and the question attributes. Here we call get_column_map:

column_map = qualtables::get_column_map(survey_id)

column_map |>
  dplyr::select(column_exported, column_harmonized, choice, question_id) |>
  head()

The first column of this data frame is the columns of the data set that qualtrics gives when responses are exported. Some of the variables are attributes of that column such as the associated choice, subq_number, and column_number. Other variables, like column_harmonized & question_name, are ones qualtables generates. I tried to make some question names and suffixes that could be used if you do not want to define them yourself. You an use the "dictionary style" renaming and label functions:

responses = qualtables::get_responses(survey_id)

responses = qualtables::rename_dict(
  df = responses,
  old_names = column_map[['column_exported']],
  new_names = column_map[['column_harmonized']]
)

responses = qualtables::var_lab_dict(
  df = responses,
  column_names = column_map[['column_harmonized']],
  var_labs = column_map[['variable_label']]
)

svenhalvorson/qualtreats documentation built on May 3, 2024, 9:33 a.m.